eir manipulation, intended or not intended, is considered lying.
             When someone is lying with statistics, they are unaware of or purposely ignoring
             statistical assumptions or rules and then make incorrect interpretations about what the data infers. An oftentimes cited example of violating statistical assumptions resulting in an incorrect result, from what I have found in my research, was the 1936 Literary Digest magazine presidential preference poll that predicted Alf Landon would defeat Franklin Roosevelt in a landslide. Roosevelt won the election. So, what went wrong?
             The pollsters at the Literary Digest used a biased sample. According to statistics, one can only make predictions from a random sample of the population, in this case, of all eligible voters. In a random sampling, each person in the population must have an equal and nonzero chance of being included in the sample. Potential respondents to the Literary Digest poll were readers of the magazine and people who had telephones. The poll was taken during the Great Depression when most poor people did not have phones. The sample was biased against poor people. In addition, the poll was biased against people with higher educations; they read the literary magazine. Wealthy people and those better educated tended to be Republicans; Landon's party. The "data" indicated Landon would win; the data, and those reporting them, lied.
             Closely related to the biased samples are those "lies" which are generated by using a small number of cases, or the "small n" problem. Consumers presented with a statement of the type "Seventy-five percent of citizens are satisfied with local services": should also be given information about the number of cases in the sample and whether the sample was drawn randomly. If a city manager can pick the citizens he or she wants to ask about the quality of services (biased sampling) and there are onl
             ...